In the field of image acquisition and processing, the use of a camera array to capture multiple images of a scene may enable the implementation of techniques for disparity and depth estimation, view interpolation, and three-dimensional reconstruction, for example. Such techniques may require determination of inter-image correspondences between pixel locations in respective images captured by the various cameras in the array. Image rectification may be employed in order to simplify these determinations. However, conventional image rectification techniques do not accommodate heterogeneous camera arrays, in which the resolutions, fields of view, and/or pixel sizes of the various cameras may differ from one another.